- Email: [email protected]
Graphene-based nanohybrid materials as the counter electrode for highly efficient quantum-dot-sensitized solar cells
CARBON
8 4 ( 2 0 1 5 ) 3 8 3 –3 8 9
Available at www.sciencedirect.com
ScienceDirect journal homepage: www.elsevier.com/locate/carbon
Graphene-based nanohybrid materials as the counter electrode for highly efficient quantum-dot-sensitized solar cells Van-Duong Dao a, Youngwoo Choi b, Kijung Yong b, Liudmila L. Larina
a,c
, Ho-Suk Choi
a,*
a
Department of Chemical Engineering, Chungnam National University, 220 Gung-Dong, Yuseong-Gu, Daejeon 305-764, Republic of Korea Department of Chemical Engineering, Division of Advanced Nuclear Engineering, Pohang University of Science and Technology (POSTECH), Pohang 790-784, Republic of Korea c Department of Solar Photovoltaics, Institute of Biochemical Physics of Russian Academy of Sciences, Russia b
A R T I C L E I N F O
A B S T R A C T
Article history:
Dry plasma reduction is an excellent approach for easily and uniformly immobilizing Pt, Au
Received 9 August 2014
and bimetallic AuPt nanoparticles (NPs) on a graphene nanoplatelets (GC)-coated layer
Accepted 3 December 2014
under atmospheric pressure at a low temperature and without using any toxic reductants.
Available online 9 December 2014
The NPs with an average size of about 2 nm were stably and uniformly hybridized on the surface of reduced graphene nanoplatelets (RGC) after co-reduction of metal precursor ions and GC to metal atoms and RGC, respectively. Quantum-dot-sensitized solar cells exploiting AuNP/RGC, PtNP/RGC and bimetallic AuPtNP/RGC counter electrodes (CEs) exhibited power conversion efficiencies of 2.7%, 3.0% and 4.5%, respectively. The efficiencies are comparable to that of device with a conventional Au-sputtered CE (3.6%). The effect is ascribed to high electrochemical catalytic activity and high electrical conductivity of developed nanohybrid materials. Ó 2014 Elsevier Ltd. All rights reserved.
1.
Introduction
Owing to using low band gap semiconductors for light-trapping, quantum-dot-sensitized solar cells (QDSCs) have attracted much attention as the next generation of dye-sensitized solar cells (DSCs) [1]. Except for the effect of quantum confinement of the exciton in the absorber, the performance of a QDSC can be improved by optimizing the band alignment at the TiO2/QD-sensitizer interface for efficient charge transfer, by increasing the light collection efficiency and by matching the absorption spectrum of the QD-sensitizer to the solar spectrum. These advantages of the QDSC can be realized by
* Corresponding author. E-mail address: [email protected] (H.-S. Choi). http://dx.doi.org/10.1016/j.carbon.2014.12.014 0008-6223/Ó 2014 Elsevier Ltd. All rights reserved.
controlling the size of the QD and the band gap energy of the semiconductor QD-sensitizer. Indeed, the value of the band gap of II–VI group compound semiconductors and the mixed metal compounds can be varied with the chemical composition of the film in the range between 1.1 and 3.7 eV. Currently, QDSCs are very promising devices to replace conventional DSCs because the energy conversion efficiency can reach 44% based upon theoretical prediction [2]. Nevertheless, until now, the champion-device efficiency of 8.5% was shown by solar cells based on Sb2S3 and PbS QDs at a laboratory scale [3]. The development of the QDSCs is of crucial importance despite the fact that these devices currently yield low efficiencies.
384
CARBON
8 4 ( 2 0 1 4 ) 3 8 3 –3 8 9
The conversion efficiency and production cost are key issues in the photovoltaic (PV) technologies today for commercialization. So far, the development of low-cost and scalable counter electrodes (CEs) with high redox activity toward the sulfide/polysulfide couple remains an issue in fabrication of QDSCs [4]. Pt and Au have been common materials for the CEs in the QDSCs due to high catalytic activity, high conductivity and chemical stability [4,5]. However, they are known to be an expensive material with limited reserves. Thus, their large-scale application is too expensive to compete with conventional solid-state PV devices. Attention to date has focused on noble-metal-free materials as alternative types of efficient electrodes. The metal chalcogenides (CoS, CuS, CuS/CoS, Cu2S, PbS, MoS2) [6–9], carbon-based materials [10], and graphene-based hybrid materials (TiN/CNT–graphene, Cu2S/graphene, Au/graphene, Mo-compound/CNT–graphene) [1,11–14] have been widely studied as low-cost cathode materials. However, the advantages of low-cost cathode materials must always be weighed against the drawbacks of their processing parameters, such as high processing temperature, chemical toxicity, and a long reduction time. Despite the many research studies on the subject, the development of low-cost and scalable CEs with high redox activity toward the sulfide/polysulfide couple remains one of the major challenges for QDSC technology to date. This research addresses one of the key questions in developing a CE for highly efficient QDSCs. For this purpose, we developed an experimental approach for co-reduction of metal precursor ions and graphene oxide through dry plasma reduction (DPR) at near room temperature under atmospheric pressure [15]. The process provides uniform hybridization of well-dispersed metal-NPs on the surface of reduced graphene oxide. We expect that the metal-NP/reduced graphene oxide nanohybrids will achieve high conductivity as well as high catalytic activity toward regenerating the S2 /Sx2 redox couple.
The assembly of the QDSCs was carried out as described in our previous study [11]. Briefly, the working electrode and CE were assembled into a sandwich type cell, which was sealed with a thermobonding polymer (Surlyn, DuPont) of 60 lm thickness at 120 °C for 5 min. A drop of the electrolyte composed of Na2S (0.5 M), S (2 M), KCl (0.2 M) in methanol/ water (the volume ratio of 7:3) was injected in each hole (0.8 mm) located in the back side of the working electrode. Two holes were sealed with Surlyn layer.
2.
Experimental section
2.4.
2.1.
Synthesis of nanohybrids materials and CEs
The morphologies of nanohybrid materials were characterized by TEM measurement (JEM-2100F, Joel, Japan). The chemical states of the nanohybrids were analyzed by energy-dispersive X-ray spectroscopy (EDS) (JEM-2100F, Joel, Japan). The reduction of GC to reduced GC (RGC) induced by the DPR process was systematically studied using X-ray photoelectron, Raman spectroscopy, and thermal gravimetric analysis (TGA) in our previous work [16]. The photocurrent density–voltage characteristics of the cells were measured under a simulated air mass 1.5 G solar spectrum. The constant light illumination was adjusted to 100 mW cm 2 using a national renewable energy laboratory (NREL)-certified silicon reference cell equipped with a KG-5 filter. An active area of 0.25 cm2 was accurately defined using a mask placed in front of the cell [11]. Electrochemical impedance spectrum (EIS) analysis of the counter electrodes was carried out in a symmetric cell configuration using an Ivium potentiostat. The frequency range was from 100 kHz to 100 mHz with a modulation amplitude of 5 mV at a 0 V bias voltage [15,16]. EIS spectra were fitted using the Z-view software package.
Three solutions containing 10 mM H2PtCl6ÆxH2O (P37.5% Pt basic, Sigma–Aldrich) and iso-propyl alcohol (IPA) (99.5%, Sigma–Aldrich), 10 mM HAuCl4Æ3H2O (Aldrich) in IPA and a solution in a mixed precursor of 10 mM H2PtCl6ÆxH2O in IPA and 10 mM HAuCl4Æ3H2O in IPA with a volume ratio of 1:1 were prepared. Next, 3 lL of the precursor was dropped onto a graphene nanoplatelets-grade C 750 m2/g (GC) CE, which was prepared separately in a previous study [16], and the solvent was allowed to evaporate at 70 °C for 15 min. Then, the specimens were reduced using Ar plasma under atmospheric pressure at a power level of 150 W, a gas flow rate of 5 lpm, a treatment time of 15 min, and a substrate moving speed of 5 mm/s. This procedure was performed with the different precursors.
2.2.
Preparation of working electrodes
The working electrodes were prepared as described in our previous study [11]. Briefly, a 50 nm-thick ZnO buffer film
was sputtered onto FTO glass (8 X/h, Pilkington, USA) and was then immersed in an aqueous solution containing Zn(NO3)2Æ6H2O (0.01 M) and NH4OH (0.5 M) for 12 h at 95 °C. Sensibilization of ZnO nanowire (NW) electrodes with CdS was conducted using the ion-layer adsorption and reaction (SILAR) method. For this purpose, the electrodes were dipped into an aqueous solution of CdSO4 (200 mM) for 30 s, rinsed with deionized water for 30 s, dipped in an aqueous solution of Na2S (200 mM) for 30 s, and finally rinsed with de-ionized water for 30 s. Then, CdSe QDs were deposited in situ on CdS/ZnO NWs using chemical bath deposition from an aqueous solution containing of 2.5 mM of Cd(CH3COO)2 (as Cd ion source), 2.5 mM of Na2SeSO3 (as Se ion source), and 45 mM of NH4OH (as complexing agent) for 3 h at 95 °C. The quantum dot depositions were carried out in a home-built bath equipped with a Teflon-bladed electric stirrer and a Teflon substrate holder. Bath temperature was adjusted to the desired temperature level using a thermostat. To achieve a suitable loading of CdSe on the CdS/ZnO-NWs, the procedure was repeated three times. The experimental details of the fabrication of working electrodes are described in previous study [11].
2.3.
Assembly of QDSCs
Characterization and measurements
CARBON
8 4 (2 0 1 4) 3 8 3–38 9
3.
Results and discussion
3.1.
Morphology and structure of nanohybrid materials
The top-view of Fig. 1 shows a schematic representation of hybridization of Pt, Au and bimetallic AuPt NPs on RGC using DPR. RGC-coated electrode was prepared first as described in our previous publication [16]. Next, 3 lL of precursor solution was dropped onto RGC-coated electrode, and the solvent was allowed to evaporate at 70 °C for 15 min. Finally, the specimens were reduced though DPR. Fig. 1a illustrates the TEM image of PtNP on the surface of RGC (PtNP/RGC). Pt NPs were clearly visible on the RGC surface without any agglomeration. The average particle size estimated from this image was around 2 nm. A high-resolution TEM image (inset of Fig. 1a) ˚, shows clear lattice fringes, with a lattice spacing of 2.26 A corresponding to the {1 1 1} plane (JCPDS 04-0802). The results are very close to the results obtained previously [16]. The TEM image of the AuNPs immobilized on the RGC film (AuNP/RGC) is given in Fig. 1b. The AuNPs with an average size of 2.5 nm are very well-dispersed on the RGC surface. As can be seen at an enlarged scale (inset of Fig. 1b), the estimated lattice ˚ , which coincides well with Au {1 1 1} (JCPDS spacing is 2.41 A 04-0784). Immobilization of the AuNPs and PtNPs on the surface of the RGC film was confirmed by EDS, which was measured during the TEM observation (Supporting Information, Figs. S1 and S2). Note that the AuNP size was slightly larger than the size of 2 nm of PtNP/RGC. This phenomena could be explained by the higher redox potential of Au compared with that of Pt. The effect of the substrate nature on the morphology of nanohybrid materials was deduced from a comparison between two nanohybrid structures formed by DPR: AuNP/RGC and AuNP/FTO glass. We found that the size of AuNPs formed on the surface of FTO glass [17] was higher than that of AuNPs formed on RGC. This finding can be explained by the difference in the electrostatic energy, which was defined by the nature of the substrates [18]. Furthermore, the NP size was also strongly affected by the nature of the reducing agents. Indeed, nanohybrid AuNP/RGC prepared by electrophoretic deposition exhibited the AuNP size in a range of 60–70 nm [13].
385
A TEM image of the bimetallic AuPt NPs decorated on the surface of RGC film (AuPtNP/RGC) is presented in Fig. 1c. The average size of the bimetallic AuPt NPs is less than 2 nm. There was no agglomeration of bimetallic NPs on the RGC surface. The inset of Fig. 1c shows clear lattice fringes ˚ , which corresponds to the with a lattice spacing of 2.37 A {1 1 1} plane of a face-centered cubic Au–Pt alloy [19]. EDS (Supporting Information, Fig. S3), which was measured during the TEM observation, identified the AuPtNPs immobilized on the surface of the RGC film. The formation of bimetallic AuPtNP on RGC film is similar to that we reported previously in details [17].
3.2. Catalytic activity of AuNP/RGC, PtNP/RGC and bimetallic AuPtNP/RGC electrodes The effect of CE on the QDSC performance is assigned to its properties related to electrical conductivity and electrocatalytic activity for the sulfide/polysulfide redox couple [11]. Therefore, the EIS analysis was conducted to get insights into the effects of the electrochemical characteristics of the CEs related to the material nature and morphology on the device performance. Nyquist plots obtained from various symmetric cells assembled by two identical CEs are presented in Fig. 2. The spectra were fitted with an extended equivalent circuit for carbon electrodes [20] using the Z-view software, which is presented in the top image of Fig. 2a. Nyquist plots are shown in Fig. 2b on an expanded scale in the dashed line square in Fig. 2a. Three semicircles were recorded in the extended equivalent circuit. The first one was described at a high frequency semicircle for electron transport resistance in the graphene layer (Rtrns) and constant phase element in the graphene layer (CPEtrap). The second one, a middle semicircle, is attributed to the charge transfer resistance at the electrode/electrolyte interface (Rct) and constant phase element at the electrode/electrolyte interface (CPEdl). The last one, a low frequency semicircle, is attributed to the Warburg impedance (W) [21]. According to this model, both the Rtrns and CPEtrap would be ignored for an Au-sputtered electrode due to a thin film structure and higher conductivity compared with carbon film [20]. The extended equivalent circuit also indicates the lower value of CPEtrap compared with CPEdl
Fig. 1 – The schematic representation of hybridization of Pt, Au and bimetallic AuPt NPs on RGC using dry plasma reduction (top-view) and TEM images of (a) PtNP/RGC, (b) AuNP/RGC, (c) bimetallic AuPt-NP/RGC (bottom-view). Insets: high-resolution TEM images of each sample. (A color version of this figure can be viewed online.)
386
CARBON
8 4 ( 2 0 1 4 ) 3 8 3 –3 8 9
Among them, CE based on bimetallic AuPtNP/RGC nanohybrid exhibits the highest conductivity, although all three electrodes with metal NPs/RGC structure reveal almost similar values of conductivity. Two factors contribute to the improvement of conductivity for nanohybrid materials compared to GC. First, the increase in conductivity is related to the reduced structural damage on the basal plane of the graphene [19] because of the loss of oxygen-containing functional groups upon reduction of GC [15]. Indeed, the oxygen-containing functional groups form sp3 bonds with carbon atoms in the basal plane. Thus, the degree of clustering of the sp2 is increased along with sp2 bond restoration during the reduction process. Another reason for the increasing conductivity of nanohybrid films is the synergistic effect of graphene with the deposited NPs [22]. The oxygen-containing functional groups play the role of nucleation centers on the surface of RGC for metal NPs [23]. NPs hybridized on the surface of the RGC CE provide a conductive network in the sp3 phase for charge carriers delocalized in the sp2 matrix (Supporting Information, Fig. S4). As a result, the electron conductivity in the metal NPs/RGC structure increases. We now turn to the experimental data related to the charge-transfer resistance for all CEs under study (Table 1). We can conclude that our experimental data on conductivity display the synergistic effect of graphene with the deposited NPs. Given that the Rct at the interface between the bimetallic AuPtNP/RGC CE and the electrolyte is the minimum value among the CEs because the synergistic effect of graphene with the AuPtNPs was stronger than the effects of graphene with the AuNPs and PtNPs. As can be seen in Table 1, the Rct of the bimetallic AuPtNP/RGC electrode was as small as 34.25 X, while that of the Au-sputtered, GC, AuNP/RGC and PtNP/RGC electrodes were 57.63, 148.7, 61.24, and 58.96 X, respectively. We made a comparison between the AuNP/ RGC, PtNP/RGC and bimetallic AuPtNP/RGC electrodes. Based on the experimental results, we can now conclude that the loading of the bimetallic NPs in the combination of the Au and Pt on GC leads to an obvious improvement in the electrocatalytic activity of the obtained CE over those shown by CEs constructed by the loading of only AuNPs or PtNPs. Furthermore, the electrocatalytic activity of the developed CE was comparable with that of the conventional Au-sputtered CE. To explain the obtained experimental results, the synergistic activity of bimetallic AuPtNP catalysts [24] and the synergistic effect of graphene with deposited bimetallic could be taken in consideration. It is well known that the exchange current density, Jo, is inversely proportional to the Rct value
Fig. 2 – (a) Nyquist plots of the dummy cells with two identical counter electrodes. The top image shows the equivalent circuit diagram used to fit EIS spectra, (b) high resolution of the dashed line square in (a). (A color version of this figure can be viewed online.)
because the trap sites only partially occupy the surface. Simulated data of EIS spectra calculated by equivalent circuits are listed in Table 1. As can be seen in Table 1, Rtrns values of AuNP/RGC, PtNP/ RGC and bimetallic AuPtNP/RGC electrodes were estimated as 2.64, 2.91 and 0.95 X, respectively, while Rtrns of the GC electrode was as high as 159.1 X (Fig. 2a and Table 1). Since the Rtrns is inversely proportional to electrode conductivity [20], all nanohybrid electrodes must have the conductivity by two orders of magnitude higher than that of the GC electrode.
Table 1 – Simulated data of EIS spectra calculated by equivalent circuits. Counter electrode
Au sputtered GC AuNP/RGC PtNP/RGC AuPtNP/RGC
Rh (X)
2.14 5.26 5.26 5.21 5.56
Rtrns (X)
– 159.1 2.64 2.91 0.95
Rct (X)
57.63 148.7 61.24 58.96 34.25
W R (X)
T
P
11.71 24.19 21.28 14.62 8.94
2.23 4.22 4.21 3.92 3.41
0.5 0.5 0.5 0.5 0.5
CPEdl-T (lF)
CPEdl-P
CPEtrap-T (lF)
CPEtrap-P
160 240 1200 920 2000
0.85 0.89 0.93 0.92 0.92
– 65.2 89.3 120 200
– 0.77 0.80 0.77 0.80
CARBON
8 4 (2 0 1 4) 3 8 3–38 9
387
Fig. 3 – QDSC with a CdSe/CdS/ZnO-NW photoelectrode and a graphene-based hybrid material CE. (A color version of this figure can be viewed online.)
(Jo = RT/nFRct) [21]. The high Jo value suggests the high fill factor (FF) and short-circuit current density (Jsc) of the QDSC [25]. The CPEdl data presented in Table 1 also confirmed that the active surface area of the bimetallic AuPtNP/RGC CE is the highest among other CEs under study. Furthermore, CPEdl data identified a larger active surface area in the bimetallic AuPtNP/RGC CE than that of the Au-sputtered CE. The diffusion impedance parameters of the electrodes are listed in Table 1. The diffusion impedance for the bimetallic AuPtNP/ RGC CE is 8.94 X. This value is smaller than the diffusion impedance of 11.71 X for the state-of-the-art CE. Note that the diffusion impedance recorded for the bimetallic CE is remarkably smaller than those of the GC, AuNP/RGC and PtNP/RGC CEs. In general, the lowest diffusion impedance indicates the easiest access of polysulfide ions to the CE. We found the lowest diffusion impedance of the bimetallic AuPtNP/RGC CE, which meant the highest conductivity as well as the best electrocatalytic activity. It is well known that a decrease of the total internal resistance related to Rct, Rtrns and diffusion impedance results in an increase in the FF [26] and Jsc [27] of the QDSC.
3.3.
under one-Sun illumination from a Sun 3000 solar simulator composed of 1000-W mercury-based Xe arc lamps and AM 1.5-G filters. Fig. 4 shows the J–V characteristics of these devices, and the device PV parameters are summarized in Table 2. We found that the QDSC employing the GC CE exhibited the smallest energy conversion efficiency, g, value of 1.5% (Fig. 4 and Table 2). The loading of NPs on GC CEs led to an improved QDSC efficiency of 2.7% for AuNP/RGC CE and 3.0% for PtNP/RGC CE. However, the efficiencies were still lower than that of the 3.6% observed for the state-of-the-art device. Two g values for QDSCs based on AuNP/RGC and PtNP/RGC CEs are slightly higher than the efficiency of 2.6% for the QDSC assembled with the Pt-sputtered CE, which we obtained in our previous study [11]. However, the most important result of our study is the highest g value of 4.5%, which was recorded for the device employing the bimetallic AuPtNP/RGC CE. In our previous study, we compared the PV
Photocurrent–voltage characteristics of QDSCs
To correlate the properties of the metal-NP/RGC nanohybrid material with the PV characteristics of devices exploiting nanohybrids as CEs, PV experiments were conducted. For this purpose, we assembled a set of QDSCs with GC, AuNP/RGC, PtNP/RGC and AuPtNP/RGC CEs. The state-of-the-art QDSC was assembled for reference. The photoanodes were fabricated by sequential deposition of low band gap semiconductors, such as CdS and CdSe, within the ZnO nanowire array as described in Fig. 3. To maximize the light harvesting efficiency of the solar cell, we used a combination of two QD sensitizers with band gap energies of 2.35 and 1.77 eV. Such a combination of semiconductors provided the optimum match of their absorption spectra with solar spectra. The photocurrent–voltage (J–V) characteristics of the QDSCs were recorded
Fig. 4 – Characteristic current density–voltage curves of QDSCs measured under standard conditions with different CEs. (A color version of this figure can be viewed online.)
388
CARBON
8 4 ( 2 0 1 4 ) 3 8 3 –3 8 9
Table 2 – The J–V characteristics of QDSCs fabricated with different CEs. Counter electrode Jsc (mA cm 2) Voc (mV) FF (%) g (%) Au sputtered GC AuNP/RGC PtNP/RGC AuPtNP/RGC
13.2 7.2 10.0 15.1 15.2
702.0 702.3 687.2 648.1 720.3
39.0 30.5 39.9 30.7 40.9
3.6 1.5 2.7 3.0 4.5
parameters of QDSC with AuPt alloy CE directly prepared on FTO glass by DPR with those of QDSC with Au-sputtered CE [17] and found that the former showed a low efficiency of 2.4% in comparison with the efficiency of 3.6% for the latter. The PV results as shown in Table 2 correlate well with the EIS analysis data. As mentioned above, the effect of the CE on PV parameters of the QDSC is assigned to its properties related to electrical conductivity and electrocatalytic activity for the sulfide/polysulfide redox couple [11]. The obtained experimental results prove that bimetallic AuPtNP/RGC may potentially replace the conventional noble metal CE, such as Au, Pt, etc. Both high electrical conductivity and electrocatalytic activity of nanohybrid bimetallic AuPtNP/RGC CEs are thought to be the primary factors that impact the increase of Jsc, FF, and Voc.
4.
Conclusion
Three types of ultrafine crystalline AuNPs, PtNPs, and bimetallic AuPtNPs were successfully hybridized on the GC though DPR at a low temperature. TEM results showed NPs with an average size of about 2 nm dispersed on the surface of the GC. The developed CE provided an active surface area that was sufficient for high catalytic activity. The QDSC employing the AuNP/RGC, PtNP/RGC and AuPtNP/RGC CEs reached efficiencies of 2.7%, 3.0% and 4.5%, respectively. The obtained efficiencies could be comparable to the efficiency of 3.6% in the QDSC with the state-of-the-art CE. The results could be attributed to the low charge-transfer resistance of the developed nanohybrid material electrodes.
Acknowledgements This research was supported by the NRF-RFBR Joint Research Program (NRF-2013K2A1A7076282), by the NRF grant (NRF-2014006994), by the Korean Brain Pool Program 2013 (131S-6-3-0538), and by the National Research Foundation Postdoctoral Fellowship Program 2014 (2014-1142-01) through the National Research Foundation of South Korea.
Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.carbon. 2014.12.014.
R E F E R E N C E S
[1] Radich JG, Dwyer R, Kamat PV. Cu2S reduced graphene oxide composite for high-efficiency quantum dot solar cells. Overcoming the redox limitations of S2 /S2n at the counter electrode. J Phys Chem Lett 2011;2:2453. [2] Hanna MC, Nozic AJJ. Solar conversion efficiency of photovoltaic and photoelectrolysis cells with carrier multiplication absorbers. J Appl Phys 2006;100: 074510. [3] Maraghechi P, Labelle AJ, Kirmani AR, Lan X, Adachi MM, Thon SM, et al. The donor-supply electrode enhances performance in colloidal quantum dot solar cells. ACS Nano 2013;7:6111. [4] Chakrapani V, Baker D, Kamat PV. Understanding the role of the sulfide redox couple (S2 /Sn2 ) in quantum dot-sensitized solar cells. J Am Chem Soc 2011;133:9607. [5] Lang Y, Lo YS. Highly efficiency quantum-dot-sensitized solar cell based on co-sensitization of CdS/CdSe. Adv Funct Mater 2009;19:604. [6] Yang Z, Chen CY, Liu CW, Li CL, Chang HT. Quantum dotsensitized solar cells featuring CuS/CoS electrodes provide 4.1% efficiency. Adv Energy Mater 2011;1:259. [7] Yang Z, Chen CY, Liu CW, Li CL, Chang HT. Electrocatalytic sulfur electrodes for CdS/CdSe quantum dot-sensitized solar cells. Chem Commun 2010;46:5485. [8] Lee JW, Son DY, Ahn TK, Shin HW, Kim IY, Hwang SJ, et al. Quantum-dot-sensitized solar cell with unprecedentedly high photocurrent. Sci Rep 2013;3:1. [9] Finn ST, Macdonald JE. Petaled molybdenum disulfide surfaces: facile synthesis of a superior cathode for QDSSCs. Adv Energy Mater 2014;4:1400495. [10] Li DM, Cheng LY, Zhang YD, Zhang QX, Huang XM, Luo YH, et al. Development of Cu2S/carbon composite electrode for CdS/CdSe quantum dot sensitized solar cell modules. Sol Energy Mater Sol Cells 2014;120:454. [11] Seol M, Kim H, Tak Y, Yong K. Novel nanowire array based highly efficient quantum dot sensitized solar cell. Chem Commun 2010;46:5521. [12] Ye M, Chen C, Zhang N, Wen X, Guo W, Lin C. Quantum-dotsensitized solar cell employing hierarchical Cu2S microsphere wrapped by reduced graphene oxide nanosheets as effective counter electrodes. Adv Energy Mater 2014;4:1301564. [13] Zhu G, Pan L, Sun H, Liu X, Lv T, Lu T, et al. Electrophoretic deposition of a reduced graphene–Au nanoparticle composite film as counter electrode for CdS quantum dotsensitized solar cells. ChemPhysChem 2012;13:769. [14] Soel M, Youn DH, Kim JY, Jang JW, Choi M, Lee JS, et al. Mocompound/CNT–graphene composites as efficient catalytic electrodes for quantum-dot-sensitized solar cells. Adv Energy Mater 2013;4:1300775. [15] Dao VD, Tran CQ, Ko SH, Choi HS. Dry plasma reduction to synthesize supported platinum nanoparticles for flexible dye-sensitized solar cells. J Mater Chem A 2013;1:4436. [16] Dao VD, Larina LL, Suh H, Hong K, Lee JK, Choi HS. Optimum strategy for designing a graphene-based counter electrode for dye-sensitized solar cells. Carbon 2014;77:980. [17] Dao VD, Choi W, Yong K, Larina LL, Shevaleevskiy O, Choi HS. A facile synthesis of bimetallic AuPt nanoparticles as a new transparent counter electrode for quantum-dot-sensitized solar cells. J Power Sources 2015;274:831. [18] Luo Z, Somers LA, Dan Y, Ly T, Kybert NJ, Mele EJ, et al. Sizeselective nanoparticle growth on few-layer graphene films. Nano Lett 2010;10:777.
CARBON
8 4 (2 0 1 4) 3 8 3–38 9
[19] Xu J, Zhao T, Liang Z, Zhu L. Facile preparation of AuPt alloy nanoparticles from organometallic complex precursor. Chem Mater 2008;20:1688. [20] Kwon W, Kim JM, Rhee SW. A new equivalent circuit model for porous carbon electrodes in charge transfer reaction of iodide/triiodide redox couples. Electrochim Acta 2012;68:110. [21] Lan Z, Wu J, Lin J, Huang M. Morphology controllable fabrication of Pt counter electrodes for highly efficient dyesensitized solar cells. J Mater Chem 2012;22:3984. [22] Huang LJ, Wang YX, Tang JG, Wang HM, Wang HB, Qiu JX, et al. Synthesis of graphene/metal nanocomposite film with good dispersibility via solvothermal method. Int J Electrochem Sci 2012;7:11068. [23] Goncalves G, Marques PAAP, Granadeiro CM, Nogueira HIS, Singh MK, Gracio J. Surface modification of graphene nanosheets with gold nanoparticles: the role of oxygen
[24]
[25]
[26]
[27]
389
moieties at graphene surface on gold nucleation and growth. Chem Mater 2009;21:4796. Mott D, Luo J, Njoki PN, Lin Y, Wang L, Zhong CJ. Synergistic activity of gold-platinum alloy nanoparticle catalysts. Catal Today 2007;122:378. Cha SI, Koo BK, Seo SH, Dong Y, Lee DY. Pt-free transparent counter electrodes for dye-sensitized solar cells prepared from carbon nanotube micro-balls. J Mater Chem 2010;20:659. Hod I, Tachan Z, Shalom M, Zaban A. Internal photoreference electrode: a powerful characterization method for photoelectrochemical quantum dot sensitized solar cells. J Phys Chem Lett 2011;2:1032. Dao VD, Kim SH, Choi HS, Kim JH, Park HO, Lee JK. Efficiency enhancement of dye-sensitized solar cell using Pt hollow sphere counter electrode. J Phys Chem C 2011;115:25529.
8 4 ( 2 0 1 5 ) 3 8 3 –3 8 9
Available at www.sciencedirect.com
ScienceDirect journal homepage: www.elsevier.com/locate/carbon
Graphene-based nanohybrid materials as the counter electrode for highly efficient quantum-dot-sensitized solar cells Van-Duong Dao a, Youngwoo Choi b, Kijung Yong b, Liudmila L. Larina
a,c
, Ho-Suk Choi
a,*
a
Department of Chemical Engineering, Chungnam National University, 220 Gung-Dong, Yuseong-Gu, Daejeon 305-764, Republic of Korea Department of Chemical Engineering, Division of Advanced Nuclear Engineering, Pohang University of Science and Technology (POSTECH), Pohang 790-784, Republic of Korea c Department of Solar Photovoltaics, Institute of Biochemical Physics of Russian Academy of Sciences, Russia b
A R T I C L E I N F O
A B S T R A C T
Article history:
Dry plasma reduction is an excellent approach for easily and uniformly immobilizing Pt, Au
Received 9 August 2014
and bimetallic AuPt nanoparticles (NPs) on a graphene nanoplatelets (GC)-coated layer
Accepted 3 December 2014
under atmospheric pressure at a low temperature and without using any toxic reductants.
Available online 9 December 2014
The NPs with an average size of about 2 nm were stably and uniformly hybridized on the surface of reduced graphene nanoplatelets (RGC) after co-reduction of metal precursor ions and GC to metal atoms and RGC, respectively. Quantum-dot-sensitized solar cells exploiting AuNP/RGC, PtNP/RGC and bimetallic AuPtNP/RGC counter electrodes (CEs) exhibited power conversion efficiencies of 2.7%, 3.0% and 4.5%, respectively. The efficiencies are comparable to that of device with a conventional Au-sputtered CE (3.6%). The effect is ascribed to high electrochemical catalytic activity and high electrical conductivity of developed nanohybrid materials. Ó 2014 Elsevier Ltd. All rights reserved.
1.
Introduction
Owing to using low band gap semiconductors for light-trapping, quantum-dot-sensitized solar cells (QDSCs) have attracted much attention as the next generation of dye-sensitized solar cells (DSCs) [1]. Except for the effect of quantum confinement of the exciton in the absorber, the performance of a QDSC can be improved by optimizing the band alignment at the TiO2/QD-sensitizer interface for efficient charge transfer, by increasing the light collection efficiency and by matching the absorption spectrum of the QD-sensitizer to the solar spectrum. These advantages of the QDSC can be realized by
* Corresponding author. E-mail address: [email protected] (H.-S. Choi). http://dx.doi.org/10.1016/j.carbon.2014.12.014 0008-6223/Ó 2014 Elsevier Ltd. All rights reserved.
controlling the size of the QD and the band gap energy of the semiconductor QD-sensitizer. Indeed, the value of the band gap of II–VI group compound semiconductors and the mixed metal compounds can be varied with the chemical composition of the film in the range between 1.1 and 3.7 eV. Currently, QDSCs are very promising devices to replace conventional DSCs because the energy conversion efficiency can reach 44% based upon theoretical prediction [2]. Nevertheless, until now, the champion-device efficiency of 8.5% was shown by solar cells based on Sb2S3 and PbS QDs at a laboratory scale [3]. The development of the QDSCs is of crucial importance despite the fact that these devices currently yield low efficiencies.
384
CARBON
8 4 ( 2 0 1 4 ) 3 8 3 –3 8 9
The conversion efficiency and production cost are key issues in the photovoltaic (PV) technologies today for commercialization. So far, the development of low-cost and scalable counter electrodes (CEs) with high redox activity toward the sulfide/polysulfide couple remains an issue in fabrication of QDSCs [4]. Pt and Au have been common materials for the CEs in the QDSCs due to high catalytic activity, high conductivity and chemical stability [4,5]. However, they are known to be an expensive material with limited reserves. Thus, their large-scale application is too expensive to compete with conventional solid-state PV devices. Attention to date has focused on noble-metal-free materials as alternative types of efficient electrodes. The metal chalcogenides (CoS, CuS, CuS/CoS, Cu2S, PbS, MoS2) [6–9], carbon-based materials [10], and graphene-based hybrid materials (TiN/CNT–graphene, Cu2S/graphene, Au/graphene, Mo-compound/CNT–graphene) [1,11–14] have been widely studied as low-cost cathode materials. However, the advantages of low-cost cathode materials must always be weighed against the drawbacks of their processing parameters, such as high processing temperature, chemical toxicity, and a long reduction time. Despite the many research studies on the subject, the development of low-cost and scalable CEs with high redox activity toward the sulfide/polysulfide couple remains one of the major challenges for QDSC technology to date. This research addresses one of the key questions in developing a CE for highly efficient QDSCs. For this purpose, we developed an experimental approach for co-reduction of metal precursor ions and graphene oxide through dry plasma reduction (DPR) at near room temperature under atmospheric pressure [15]. The process provides uniform hybridization of well-dispersed metal-NPs on the surface of reduced graphene oxide. We expect that the metal-NP/reduced graphene oxide nanohybrids will achieve high conductivity as well as high catalytic activity toward regenerating the S2 /Sx2 redox couple.
The assembly of the QDSCs was carried out as described in our previous study [11]. Briefly, the working electrode and CE were assembled into a sandwich type cell, which was sealed with a thermobonding polymer (Surlyn, DuPont) of 60 lm thickness at 120 °C for 5 min. A drop of the electrolyte composed of Na2S (0.5 M), S (2 M), KCl (0.2 M) in methanol/ water (the volume ratio of 7:3) was injected in each hole (0.8 mm) located in the back side of the working electrode. Two holes were sealed with Surlyn layer.
2.
Experimental section
2.4.
2.1.
Synthesis of nanohybrids materials and CEs
The morphologies of nanohybrid materials were characterized by TEM measurement (JEM-2100F, Joel, Japan). The chemical states of the nanohybrids were analyzed by energy-dispersive X-ray spectroscopy (EDS) (JEM-2100F, Joel, Japan). The reduction of GC to reduced GC (RGC) induced by the DPR process was systematically studied using X-ray photoelectron, Raman spectroscopy, and thermal gravimetric analysis (TGA) in our previous work [16]. The photocurrent density–voltage characteristics of the cells were measured under a simulated air mass 1.5 G solar spectrum. The constant light illumination was adjusted to 100 mW cm 2 using a national renewable energy laboratory (NREL)-certified silicon reference cell equipped with a KG-5 filter. An active area of 0.25 cm2 was accurately defined using a mask placed in front of the cell [11]. Electrochemical impedance spectrum (EIS) analysis of the counter electrodes was carried out in a symmetric cell configuration using an Ivium potentiostat. The frequency range was from 100 kHz to 100 mHz with a modulation amplitude of 5 mV at a 0 V bias voltage [15,16]. EIS spectra were fitted using the Z-view software package.
Three solutions containing 10 mM H2PtCl6ÆxH2O (P37.5% Pt basic, Sigma–Aldrich) and iso-propyl alcohol (IPA) (99.5%, Sigma–Aldrich), 10 mM HAuCl4Æ3H2O (Aldrich) in IPA and a solution in a mixed precursor of 10 mM H2PtCl6ÆxH2O in IPA and 10 mM HAuCl4Æ3H2O in IPA with a volume ratio of 1:1 were prepared. Next, 3 lL of the precursor was dropped onto a graphene nanoplatelets-grade C 750 m2/g (GC) CE, which was prepared separately in a previous study [16], and the solvent was allowed to evaporate at 70 °C for 15 min. Then, the specimens were reduced using Ar plasma under atmospheric pressure at a power level of 150 W, a gas flow rate of 5 lpm, a treatment time of 15 min, and a substrate moving speed of 5 mm/s. This procedure was performed with the different precursors.
2.2.
Preparation of working electrodes
The working electrodes were prepared as described in our previous study [11]. Briefly, a 50 nm-thick ZnO buffer film
was sputtered onto FTO glass (8 X/h, Pilkington, USA) and was then immersed in an aqueous solution containing Zn(NO3)2Æ6H2O (0.01 M) and NH4OH (0.5 M) for 12 h at 95 °C. Sensibilization of ZnO nanowire (NW) electrodes with CdS was conducted using the ion-layer adsorption and reaction (SILAR) method. For this purpose, the electrodes were dipped into an aqueous solution of CdSO4 (200 mM) for 30 s, rinsed with deionized water for 30 s, dipped in an aqueous solution of Na2S (200 mM) for 30 s, and finally rinsed with de-ionized water for 30 s. Then, CdSe QDs were deposited in situ on CdS/ZnO NWs using chemical bath deposition from an aqueous solution containing of 2.5 mM of Cd(CH3COO)2 (as Cd ion source), 2.5 mM of Na2SeSO3 (as Se ion source), and 45 mM of NH4OH (as complexing agent) for 3 h at 95 °C. The quantum dot depositions were carried out in a home-built bath equipped with a Teflon-bladed electric stirrer and a Teflon substrate holder. Bath temperature was adjusted to the desired temperature level using a thermostat. To achieve a suitable loading of CdSe on the CdS/ZnO-NWs, the procedure was repeated three times. The experimental details of the fabrication of working electrodes are described in previous study [11].
2.3.
Assembly of QDSCs
Characterization and measurements
CARBON
8 4 (2 0 1 4) 3 8 3–38 9
3.
Results and discussion
3.1.
Morphology and structure of nanohybrid materials
The top-view of Fig. 1 shows a schematic representation of hybridization of Pt, Au and bimetallic AuPt NPs on RGC using DPR. RGC-coated electrode was prepared first as described in our previous publication [16]. Next, 3 lL of precursor solution was dropped onto RGC-coated electrode, and the solvent was allowed to evaporate at 70 °C for 15 min. Finally, the specimens were reduced though DPR. Fig. 1a illustrates the TEM image of PtNP on the surface of RGC (PtNP/RGC). Pt NPs were clearly visible on the RGC surface without any agglomeration. The average particle size estimated from this image was around 2 nm. A high-resolution TEM image (inset of Fig. 1a) ˚, shows clear lattice fringes, with a lattice spacing of 2.26 A corresponding to the {1 1 1} plane (JCPDS 04-0802). The results are very close to the results obtained previously [16]. The TEM image of the AuNPs immobilized on the RGC film (AuNP/RGC) is given in Fig. 1b. The AuNPs with an average size of 2.5 nm are very well-dispersed on the RGC surface. As can be seen at an enlarged scale (inset of Fig. 1b), the estimated lattice ˚ , which coincides well with Au {1 1 1} (JCPDS spacing is 2.41 A 04-0784). Immobilization of the AuNPs and PtNPs on the surface of the RGC film was confirmed by EDS, which was measured during the TEM observation (Supporting Information, Figs. S1 and S2). Note that the AuNP size was slightly larger than the size of 2 nm of PtNP/RGC. This phenomena could be explained by the higher redox potential of Au compared with that of Pt. The effect of the substrate nature on the morphology of nanohybrid materials was deduced from a comparison between two nanohybrid structures formed by DPR: AuNP/RGC and AuNP/FTO glass. We found that the size of AuNPs formed on the surface of FTO glass [17] was higher than that of AuNPs formed on RGC. This finding can be explained by the difference in the electrostatic energy, which was defined by the nature of the substrates [18]. Furthermore, the NP size was also strongly affected by the nature of the reducing agents. Indeed, nanohybrid AuNP/RGC prepared by electrophoretic deposition exhibited the AuNP size in a range of 60–70 nm [13].
385
A TEM image of the bimetallic AuPt NPs decorated on the surface of RGC film (AuPtNP/RGC) is presented in Fig. 1c. The average size of the bimetallic AuPt NPs is less than 2 nm. There was no agglomeration of bimetallic NPs on the RGC surface. The inset of Fig. 1c shows clear lattice fringes ˚ , which corresponds to the with a lattice spacing of 2.37 A {1 1 1} plane of a face-centered cubic Au–Pt alloy [19]. EDS (Supporting Information, Fig. S3), which was measured during the TEM observation, identified the AuPtNPs immobilized on the surface of the RGC film. The formation of bimetallic AuPtNP on RGC film is similar to that we reported previously in details [17].
3.2. Catalytic activity of AuNP/RGC, PtNP/RGC and bimetallic AuPtNP/RGC electrodes The effect of CE on the QDSC performance is assigned to its properties related to electrical conductivity and electrocatalytic activity for the sulfide/polysulfide redox couple [11]. Therefore, the EIS analysis was conducted to get insights into the effects of the electrochemical characteristics of the CEs related to the material nature and morphology on the device performance. Nyquist plots obtained from various symmetric cells assembled by two identical CEs are presented in Fig. 2. The spectra were fitted with an extended equivalent circuit for carbon electrodes [20] using the Z-view software, which is presented in the top image of Fig. 2a. Nyquist plots are shown in Fig. 2b on an expanded scale in the dashed line square in Fig. 2a. Three semicircles were recorded in the extended equivalent circuit. The first one was described at a high frequency semicircle for electron transport resistance in the graphene layer (Rtrns) and constant phase element in the graphene layer (CPEtrap). The second one, a middle semicircle, is attributed to the charge transfer resistance at the electrode/electrolyte interface (Rct) and constant phase element at the electrode/electrolyte interface (CPEdl). The last one, a low frequency semicircle, is attributed to the Warburg impedance (W) [21]. According to this model, both the Rtrns and CPEtrap would be ignored for an Au-sputtered electrode due to a thin film structure and higher conductivity compared with carbon film [20]. The extended equivalent circuit also indicates the lower value of CPEtrap compared with CPEdl
Fig. 1 – The schematic representation of hybridization of Pt, Au and bimetallic AuPt NPs on RGC using dry plasma reduction (top-view) and TEM images of (a) PtNP/RGC, (b) AuNP/RGC, (c) bimetallic AuPt-NP/RGC (bottom-view). Insets: high-resolution TEM images of each sample. (A color version of this figure can be viewed online.)
386
CARBON
8 4 ( 2 0 1 4 ) 3 8 3 –3 8 9
Among them, CE based on bimetallic AuPtNP/RGC nanohybrid exhibits the highest conductivity, although all three electrodes with metal NPs/RGC structure reveal almost similar values of conductivity. Two factors contribute to the improvement of conductivity for nanohybrid materials compared to GC. First, the increase in conductivity is related to the reduced structural damage on the basal plane of the graphene [19] because of the loss of oxygen-containing functional groups upon reduction of GC [15]. Indeed, the oxygen-containing functional groups form sp3 bonds with carbon atoms in the basal plane. Thus, the degree of clustering of the sp2 is increased along with sp2 bond restoration during the reduction process. Another reason for the increasing conductivity of nanohybrid films is the synergistic effect of graphene with the deposited NPs [22]. The oxygen-containing functional groups play the role of nucleation centers on the surface of RGC for metal NPs [23]. NPs hybridized on the surface of the RGC CE provide a conductive network in the sp3 phase for charge carriers delocalized in the sp2 matrix (Supporting Information, Fig. S4). As a result, the electron conductivity in the metal NPs/RGC structure increases. We now turn to the experimental data related to the charge-transfer resistance for all CEs under study (Table 1). We can conclude that our experimental data on conductivity display the synergistic effect of graphene with the deposited NPs. Given that the Rct at the interface between the bimetallic AuPtNP/RGC CE and the electrolyte is the minimum value among the CEs because the synergistic effect of graphene with the AuPtNPs was stronger than the effects of graphene with the AuNPs and PtNPs. As can be seen in Table 1, the Rct of the bimetallic AuPtNP/RGC electrode was as small as 34.25 X, while that of the Au-sputtered, GC, AuNP/RGC and PtNP/RGC electrodes were 57.63, 148.7, 61.24, and 58.96 X, respectively. We made a comparison between the AuNP/ RGC, PtNP/RGC and bimetallic AuPtNP/RGC electrodes. Based on the experimental results, we can now conclude that the loading of the bimetallic NPs in the combination of the Au and Pt on GC leads to an obvious improvement in the electrocatalytic activity of the obtained CE over those shown by CEs constructed by the loading of only AuNPs or PtNPs. Furthermore, the electrocatalytic activity of the developed CE was comparable with that of the conventional Au-sputtered CE. To explain the obtained experimental results, the synergistic activity of bimetallic AuPtNP catalysts [24] and the synergistic effect of graphene with deposited bimetallic could be taken in consideration. It is well known that the exchange current density, Jo, is inversely proportional to the Rct value
Fig. 2 – (a) Nyquist plots of the dummy cells with two identical counter electrodes. The top image shows the equivalent circuit diagram used to fit EIS spectra, (b) high resolution of the dashed line square in (a). (A color version of this figure can be viewed online.)
because the trap sites only partially occupy the surface. Simulated data of EIS spectra calculated by equivalent circuits are listed in Table 1. As can be seen in Table 1, Rtrns values of AuNP/RGC, PtNP/ RGC and bimetallic AuPtNP/RGC electrodes were estimated as 2.64, 2.91 and 0.95 X, respectively, while Rtrns of the GC electrode was as high as 159.1 X (Fig. 2a and Table 1). Since the Rtrns is inversely proportional to electrode conductivity [20], all nanohybrid electrodes must have the conductivity by two orders of magnitude higher than that of the GC electrode.
Table 1 – Simulated data of EIS spectra calculated by equivalent circuits. Counter electrode
Au sputtered GC AuNP/RGC PtNP/RGC AuPtNP/RGC
Rh (X)
2.14 5.26 5.26 5.21 5.56
Rtrns (X)
– 159.1 2.64 2.91 0.95
Rct (X)
57.63 148.7 61.24 58.96 34.25
W R (X)
T
P
11.71 24.19 21.28 14.62 8.94
2.23 4.22 4.21 3.92 3.41
0.5 0.5 0.5 0.5 0.5
CPEdl-T (lF)
CPEdl-P
CPEtrap-T (lF)
CPEtrap-P
160 240 1200 920 2000
0.85 0.89 0.93 0.92 0.92
– 65.2 89.3 120 200
– 0.77 0.80 0.77 0.80
CARBON
8 4 (2 0 1 4) 3 8 3–38 9
387
Fig. 3 – QDSC with a CdSe/CdS/ZnO-NW photoelectrode and a graphene-based hybrid material CE. (A color version of this figure can be viewed online.)
(Jo = RT/nFRct) [21]. The high Jo value suggests the high fill factor (FF) and short-circuit current density (Jsc) of the QDSC [25]. The CPEdl data presented in Table 1 also confirmed that the active surface area of the bimetallic AuPtNP/RGC CE is the highest among other CEs under study. Furthermore, CPEdl data identified a larger active surface area in the bimetallic AuPtNP/RGC CE than that of the Au-sputtered CE. The diffusion impedance parameters of the electrodes are listed in Table 1. The diffusion impedance for the bimetallic AuPtNP/ RGC CE is 8.94 X. This value is smaller than the diffusion impedance of 11.71 X for the state-of-the-art CE. Note that the diffusion impedance recorded for the bimetallic CE is remarkably smaller than those of the GC, AuNP/RGC and PtNP/RGC CEs. In general, the lowest diffusion impedance indicates the easiest access of polysulfide ions to the CE. We found the lowest diffusion impedance of the bimetallic AuPtNP/RGC CE, which meant the highest conductivity as well as the best electrocatalytic activity. It is well known that a decrease of the total internal resistance related to Rct, Rtrns and diffusion impedance results in an increase in the FF [26] and Jsc [27] of the QDSC.
3.3.
under one-Sun illumination from a Sun 3000 solar simulator composed of 1000-W mercury-based Xe arc lamps and AM 1.5-G filters. Fig. 4 shows the J–V characteristics of these devices, and the device PV parameters are summarized in Table 2. We found that the QDSC employing the GC CE exhibited the smallest energy conversion efficiency, g, value of 1.5% (Fig. 4 and Table 2). The loading of NPs on GC CEs led to an improved QDSC efficiency of 2.7% for AuNP/RGC CE and 3.0% for PtNP/RGC CE. However, the efficiencies were still lower than that of the 3.6% observed for the state-of-the-art device. Two g values for QDSCs based on AuNP/RGC and PtNP/RGC CEs are slightly higher than the efficiency of 2.6% for the QDSC assembled with the Pt-sputtered CE, which we obtained in our previous study [11]. However, the most important result of our study is the highest g value of 4.5%, which was recorded for the device employing the bimetallic AuPtNP/RGC CE. In our previous study, we compared the PV
Photocurrent–voltage characteristics of QDSCs
To correlate the properties of the metal-NP/RGC nanohybrid material with the PV characteristics of devices exploiting nanohybrids as CEs, PV experiments were conducted. For this purpose, we assembled a set of QDSCs with GC, AuNP/RGC, PtNP/RGC and AuPtNP/RGC CEs. The state-of-the-art QDSC was assembled for reference. The photoanodes were fabricated by sequential deposition of low band gap semiconductors, such as CdS and CdSe, within the ZnO nanowire array as described in Fig. 3. To maximize the light harvesting efficiency of the solar cell, we used a combination of two QD sensitizers with band gap energies of 2.35 and 1.77 eV. Such a combination of semiconductors provided the optimum match of their absorption spectra with solar spectra. The photocurrent–voltage (J–V) characteristics of the QDSCs were recorded
Fig. 4 – Characteristic current density–voltage curves of QDSCs measured under standard conditions with different CEs. (A color version of this figure can be viewed online.)
388
CARBON
8 4 ( 2 0 1 4 ) 3 8 3 –3 8 9
Table 2 – The J–V characteristics of QDSCs fabricated with different CEs. Counter electrode Jsc (mA cm 2) Voc (mV) FF (%) g (%) Au sputtered GC AuNP/RGC PtNP/RGC AuPtNP/RGC
13.2 7.2 10.0 15.1 15.2
702.0 702.3 687.2 648.1 720.3
39.0 30.5 39.9 30.7 40.9
3.6 1.5 2.7 3.0 4.5
parameters of QDSC with AuPt alloy CE directly prepared on FTO glass by DPR with those of QDSC with Au-sputtered CE [17] and found that the former showed a low efficiency of 2.4% in comparison with the efficiency of 3.6% for the latter. The PV results as shown in Table 2 correlate well with the EIS analysis data. As mentioned above, the effect of the CE on PV parameters of the QDSC is assigned to its properties related to electrical conductivity and electrocatalytic activity for the sulfide/polysulfide redox couple [11]. The obtained experimental results prove that bimetallic AuPtNP/RGC may potentially replace the conventional noble metal CE, such as Au, Pt, etc. Both high electrical conductivity and electrocatalytic activity of nanohybrid bimetallic AuPtNP/RGC CEs are thought to be the primary factors that impact the increase of Jsc, FF, and Voc.
4.
Conclusion
Three types of ultrafine crystalline AuNPs, PtNPs, and bimetallic AuPtNPs were successfully hybridized on the GC though DPR at a low temperature. TEM results showed NPs with an average size of about 2 nm dispersed on the surface of the GC. The developed CE provided an active surface area that was sufficient for high catalytic activity. The QDSC employing the AuNP/RGC, PtNP/RGC and AuPtNP/RGC CEs reached efficiencies of 2.7%, 3.0% and 4.5%, respectively. The obtained efficiencies could be comparable to the efficiency of 3.6% in the QDSC with the state-of-the-art CE. The results could be attributed to the low charge-transfer resistance of the developed nanohybrid material electrodes.
Acknowledgements This research was supported by the NRF-RFBR Joint Research Program (NRF-2013K2A1A7076282), by the NRF grant (NRF-2014006994), by the Korean Brain Pool Program 2013 (131S-6-3-0538), and by the National Research Foundation Postdoctoral Fellowship Program 2014 (2014-1142-01) through the National Research Foundation of South Korea.
Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.carbon. 2014.12.014.
R E F E R E N C E S
[1] Radich JG, Dwyer R, Kamat PV. Cu2S reduced graphene oxide composite for high-efficiency quantum dot solar cells. Overcoming the redox limitations of S2 /S2n at the counter electrode. J Phys Chem Lett 2011;2:2453. [2] Hanna MC, Nozic AJJ. Solar conversion efficiency of photovoltaic and photoelectrolysis cells with carrier multiplication absorbers. J Appl Phys 2006;100: 074510. [3] Maraghechi P, Labelle AJ, Kirmani AR, Lan X, Adachi MM, Thon SM, et al. The donor-supply electrode enhances performance in colloidal quantum dot solar cells. ACS Nano 2013;7:6111. [4] Chakrapani V, Baker D, Kamat PV. Understanding the role of the sulfide redox couple (S2 /Sn2 ) in quantum dot-sensitized solar cells. J Am Chem Soc 2011;133:9607. [5] Lang Y, Lo YS. Highly efficiency quantum-dot-sensitized solar cell based on co-sensitization of CdS/CdSe. Adv Funct Mater 2009;19:604. [6] Yang Z, Chen CY, Liu CW, Li CL, Chang HT. Quantum dotsensitized solar cells featuring CuS/CoS electrodes provide 4.1% efficiency. Adv Energy Mater 2011;1:259. [7] Yang Z, Chen CY, Liu CW, Li CL, Chang HT. Electrocatalytic sulfur electrodes for CdS/CdSe quantum dot-sensitized solar cells. Chem Commun 2010;46:5485. [8] Lee JW, Son DY, Ahn TK, Shin HW, Kim IY, Hwang SJ, et al. Quantum-dot-sensitized solar cell with unprecedentedly high photocurrent. Sci Rep 2013;3:1. [9] Finn ST, Macdonald JE. Petaled molybdenum disulfide surfaces: facile synthesis of a superior cathode for QDSSCs. Adv Energy Mater 2014;4:1400495. [10] Li DM, Cheng LY, Zhang YD, Zhang QX, Huang XM, Luo YH, et al. Development of Cu2S/carbon composite electrode for CdS/CdSe quantum dot sensitized solar cell modules. Sol Energy Mater Sol Cells 2014;120:454. [11] Seol M, Kim H, Tak Y, Yong K. Novel nanowire array based highly efficient quantum dot sensitized solar cell. Chem Commun 2010;46:5521. [12] Ye M, Chen C, Zhang N, Wen X, Guo W, Lin C. Quantum-dotsensitized solar cell employing hierarchical Cu2S microsphere wrapped by reduced graphene oxide nanosheets as effective counter electrodes. Adv Energy Mater 2014;4:1301564. [13] Zhu G, Pan L, Sun H, Liu X, Lv T, Lu T, et al. Electrophoretic deposition of a reduced graphene–Au nanoparticle composite film as counter electrode for CdS quantum dotsensitized solar cells. ChemPhysChem 2012;13:769. [14] Soel M, Youn DH, Kim JY, Jang JW, Choi M, Lee JS, et al. Mocompound/CNT–graphene composites as efficient catalytic electrodes for quantum-dot-sensitized solar cells. Adv Energy Mater 2013;4:1300775. [15] Dao VD, Tran CQ, Ko SH, Choi HS. Dry plasma reduction to synthesize supported platinum nanoparticles for flexible dye-sensitized solar cells. J Mater Chem A 2013;1:4436. [16] Dao VD, Larina LL, Suh H, Hong K, Lee JK, Choi HS. Optimum strategy for designing a graphene-based counter electrode for dye-sensitized solar cells. Carbon 2014;77:980. [17] Dao VD, Choi W, Yong K, Larina LL, Shevaleevskiy O, Choi HS. A facile synthesis of bimetallic AuPt nanoparticles as a new transparent counter electrode for quantum-dot-sensitized solar cells. J Power Sources 2015;274:831. [18] Luo Z, Somers LA, Dan Y, Ly T, Kybert NJ, Mele EJ, et al. Sizeselective nanoparticle growth on few-layer graphene films. Nano Lett 2010;10:777.
CARBON
8 4 (2 0 1 4) 3 8 3–38 9
[19] Xu J, Zhao T, Liang Z, Zhu L. Facile preparation of AuPt alloy nanoparticles from organometallic complex precursor. Chem Mater 2008;20:1688. [20] Kwon W, Kim JM, Rhee SW. A new equivalent circuit model for porous carbon electrodes in charge transfer reaction of iodide/triiodide redox couples. Electrochim Acta 2012;68:110. [21] Lan Z, Wu J, Lin J, Huang M. Morphology controllable fabrication of Pt counter electrodes for highly efficient dyesensitized solar cells. J Mater Chem 2012;22:3984. [22] Huang LJ, Wang YX, Tang JG, Wang HM, Wang HB, Qiu JX, et al. Synthesis of graphene/metal nanocomposite film with good dispersibility via solvothermal method. Int J Electrochem Sci 2012;7:11068. [23] Goncalves G, Marques PAAP, Granadeiro CM, Nogueira HIS, Singh MK, Gracio J. Surface modification of graphene nanosheets with gold nanoparticles: the role of oxygen
[24]
[25]
[26]
[27]
389
moieties at graphene surface on gold nucleation and growth. Chem Mater 2009;21:4796. Mott D, Luo J, Njoki PN, Lin Y, Wang L, Zhong CJ. Synergistic activity of gold-platinum alloy nanoparticle catalysts. Catal Today 2007;122:378. Cha SI, Koo BK, Seo SH, Dong Y, Lee DY. Pt-free transparent counter electrodes for dye-sensitized solar cells prepared from carbon nanotube micro-balls. J Mater Chem 2010;20:659. Hod I, Tachan Z, Shalom M, Zaban A. Internal photoreference electrode: a powerful characterization method for photoelectrochemical quantum dot sensitized solar cells. J Phys Chem Lett 2011;2:1032. Dao VD, Kim SH, Choi HS, Kim JH, Park HO, Lee JK. Efficiency enhancement of dye-sensitized solar cell using Pt hollow sphere counter electrode. J Phys Chem C 2011;115:25529.